Longer jail sentences do deter crime, but only up to a point
By S.K. | LONDON
DO CRIMINALS trade short-term gain for long-term pain? Economists have long suspected that those who commit crimes place less value on the future than law-abiding citizens. But they have mostly struggled to find hard evidence that criminals think about sentence lengths at all. A review by Steven Durlauf of the University of Wisconsin and Daniel Nagin at Carnegie Mellon University found little evidence that criminals responded to harsher sentencing, and much stronger evidence that increasing the certainty of punishment deterred crime. This matters for policy, as it suggests that locking vast numbers of people in jail is not only expensive, but useless as a deterrent.
A new working paper presented at the Royal Economic Society annual conference, by Giovanni Mastrobuoni at the University of Essex and David Rivers at the University of Western Ontario, provides new evidence that the story is a little more nuanced; criminals do value the future, just not as much as the average person. Harsher sentences work as a deterrent, but only up to a point.
The economists take advantage of an Italian policy reform in 2006, where more than a third of the Italian prison population was released. Their get-out-of-jail-free card came with one caveat: if they were caught breaking the law again, their unserved sentence would be automatically added to the new one. This was a boon for the economists, as it meant that otherwise very similar people faced gentler or harsher sentences for future crimes, depending on how much time they had left on their existing sentence (i.e. when they had committed their original crime). The economists combined this with information on reoffending rates, to see if people with harsher sentences were less likely to reoffend.
They found that the convicts did seem to respond to the harsher sentences. They estimated that previously convicted criminals discount the future at a rate of 0.74: in other words, they care about events in one year’s time around three quarters as much as events today. This compares to discount rates of around 0.95 more common in the population at large. The economists compare different groups; the highly educated are most sensitive to heftier sentences, while immigrants and drug-offenders are least sensitive. Their study is limited as it concentrates on a group of people who have already served jail time, who might be quite different to the wider population of potential criminals. For example, they are more likely to be aware of sentencing rules, and so could be more sensitive to the deterrent effects of longer sentences.
With that caveat in mind, their results have several implications for criminal justice policy. First, longer sentences might work better to deter white-collar crime than drug offences. Second, while longer sentences do have some deterrent effect, they are concentrated in the first few years. A mandatory minimum sentence of ten years with a threat of an extra ten depending on the nature of the crime may deter some crimes. But the extra ten years will be less effective as a deterrent than the threat of five years on top of a five year mandatory minimum. According to the estimates in the paper, criminals worry about the extra five years four times as much as the extra ten.
This result should be a win-win for policymakers. Shorter mandatory minimum sentences could improve the deterrent effect of (discretionary) extra jail time for worse crimes, save public money, and restore people's freedom. Policymakers in America, where 2.2 million Americans were incarcerated in 2014, seem to be coming round to this idea. Reform of the criminal justice system is one of the few causes that still has bipartisan support; the Sentencing Reform and Corrections Act of 2015, which would reduce mandatory minimum sentences for crimes including some drug offences, is currently going through Congress. And last week, the Maryland Senate approved a bill that would put more emphasis on treatment rather than punishment.
Economists have a lot to offer in this field, as they can work out which (and how much) incentives matter. But policymakers should use their results with care; although criminals may adjust their decisions depending on the severity of punishment and the chances of getting caught, their preferences may not be fixed. Understanding why criminals are short-termist could be just as important for cutting crime as knowing how to set sentences.
